Abstract: A new algorithm for multi-objective optimization problems is proposed. The antibodies in the antibody population are divided into dominated ones and non-dominated ones, which is suitable for the characteristic that one multi-objective optimization problem has a series Pareto-optimal solutions. Selecting of the non-dominated antibodies guarantees the convergence to the true Pareto-front and the convergence speed. The clonal operation implements the searching for optimal solutions in the global region and is available for getting a widely spread Pareto-front. Adopting the non-uniform mutation operation improves the searching for optimal solutions in the local region and assures the diversity of the solutions. Compared with the existing algorithms, simulation results show that the solutions obtained by the new algorithm are most widely spread, dominate those gained by the other algorithms to some extent and depress the metric S to less than 3%.

Key words: clonal selection, multi-objective optimization, non-uniform mutation, performance metrics